Brief Summary of Research Projects
Software Defects and Products Liability
-
-
Software engineering processes are inadequate to the task
of producing safe, reliable systems. They always will be. This is not a
new problem. However, software adds a new dimension to the problem because
its unique nature renders much of the traditional engineering wisdom inapplicab\
le.
Safety
constraints for artifacts of engineering processes are derived primarily
from social expectations. They have a powerful expression in the law of
products liability. Judges and lawyers currently struggle to derive reasonable
legal expectations with respect to the software artifact. Like the engineers,
they have mainly the methods and tools based on [legal and technical] models
created for traditionally engineered artifacts.
My research questions the
adequacy of the basic engineering and legal models used to address the
problem of unsafe software systems. Insights gained from this work have
important implications for both software engineering and the law.
I began my work by proposing a tool that could guide the
software engineer through the requirments of the law of negligence for
safety-critical software. This resulted in a paper
presented at SAFECOMP '96 in Vienna, Austria. I worked on this project
until it became obvious that there was something about software and its
interface with current law that made a general solution impossible.
I then investigated the ability of the software engineering and the
legal communities to reasonably define the software product and to outline
socially acceptable limits of tort responsibility for software related
personal injuries. I have a preliminary version of the main ideas published
in a UCI Technical
Report No. 99-17. I finished my dissertation entitled,
"Software as Product: The Technical Challenges
to Social Notions of Responsibility" in August of 1999.
Continuing this work, Professor Debra Richardson and I submitted a
paper entitled, "Software Control and
Strict Products Liability: A Technical Challenge to Current Legal Notions
of Responsibility" published by the International Association of
Science and
Technology for Development, LawTech 2000 conference.
Further work continues by application of the principles to the management of
software project risks in a draft paper, "Risk
Management for Safety-Critical Software: A Unique Problem on the Horizon,"
published in the "Technology Report," a publication of the Technology
Section of the Academy of Legal Studies in Business. This Journal may be
found at
www.rmi.gsu.edu/legal/technologyreport/techindex.htm
(or at its permanent home
www.alsb.org soon.)
We also begin some workflow efforts, based on the work cited above for
Safecomp, and providing automated assistance for organizations involved with
safety-critical legal risks. Our first attempt to model the situation
was given for the IASTED Software
Engineering and Applications conference in 2001, entitled, "Rethinking Software Process: The Key to Negligence Liability."
Currently, I am working on further development of the key legal principles and
their specific application to common software code defects.
Software Copyright Law and Safety
I recently thought about the economic incentives promulgated by US
Copyright Law, particularly the ability to copyright binary code while
keeping the source as a trade secret. Besides the standard Open Source
arguments for the "public good" I found that current law fosters some
compromises in public safety of software controlled machines. Here is
my recent slide presentation for the library, entitled,
"Life, Death and Copyright, Really."
Satellite Work
Watch our polysat project
We are designing and building Cubesats and a standard deployer here at
Cal Poly! Tiny satellites powered by
batteries and carrying amateur radio provide endless opportunities for
fun and learning for the software crowd. Perhaps even more importantly,
this forces our software engineers to interface with engineers from many
other disciplines to achieve a common goal. For a good overview, refer to
one of the papers recently published by our group,
"Development of the Standard CubeSat Deployer and a CubeSat Class
PicoSatellite". Some later work is included in our small satellite
conference publication of "CubeSat: The Development and
Launch Support Infrastructure for Eighteen Different Satellite Customers on
One Launch".
Software Engineering Education
We have developed a new degree program in Software Engineering
(in the approval process) and our
preliminary results for the capstone design courses were published
in
ICEE in the summer of 2001. We've since published another paper for ICEE in 2003 and further work was published for ASEE in 2004. These pdf's are for your viewing only, please refer to the publications for reprints.
Clark Savage Turner
Department of Computer Science
California Polytechnic State University
San Luis Obispo, CA 93407
csturner@calpoly.edu